Formation of functional neural circuits is critical for proper functioning of the brain. To establish the most efficient synaptic circuits, synaptic connections must be refined by neural activity during development. In this proposal, we will determine the molecules and manner by which functional circuits are established by neural activity, focusing on the limbic system (including the hippocampus and cingulate cortex), which is implicated in emotional processing, memory formation and social behavior. Using a mouse genetic system in which restricted populations of hippocampal neurons can be conditionally inactivated, we found that hippocampal axons are refined through activity-dependent competition, where active neuronal connections stay (maintained) while inactive ones leave (eliminated). We further found that a cell adhesion molecule SIRP? from postsynaptic neurons stabilizes active synapses through its presynaptic receptor CD47, serving as a Stay signal. To identify the signaling molecules that play critical roles in inactive axon elimination (Go signal), we generated a new system in which neural activity and gene expression can be conditionally controlled in vivo, using in utero electroporation. When neurotransmitter release is blocked in a subset of neurons in the cingulate cortex, their callosal projections (the major connections between the cerebral hemispheres) are eliminated during development. Using this system, we screened for signaling molecules that are upregulated in inactive neurons right before their axons start to leave and identified the Ca2+-dependent tyrosine kinase Pyk2. Inactive axons were not eliminated when a kinase-dead mutant of Pyk2 was expressed, indicating that Pyk2 activity is necessary for inactive axons to leave. We further identified that a Pyk2-interacting kinase, JAK2, is also necessary for inactive axon elimination. Consistently, Pyk2 and JAK2 are activated in inactive neurons. Finally, we found that overexpression of Pyk2 or JAK2 induces axonal elimination even when axons are active. We propose that the Pyk2-JAK2 pathway is the Go signal and serves as the determinant of axon refinement. To further characterize this pathway and to understand how the Go and Stay pathways regulate activity- dependent axon/synapse refinement, we propose to: Aim 1: Investigate the role of Pyk2 and JAK2 for axon/synapse refinement in physiological conditions. Aim 2: Analyze the electrophysiological consequences of Pyk2/JAK2 inactivation during synapse refinement using conditional KO mice. Aim 3: Examine whether the Pyk2-JAK2 pathway provides cues for microglial clearance of inactive axons in vivo. Aim 4: Investigate the interaction between the Stay (SIRP?-CD47) and Go (Pyk2-JAK2) pathways in axon/synapse refinement in vivo. Our project will molecularly delineate how neurons decide to establish functional synaptic connections in the mammalian brain. Pyk2 and JAK2 are associated with various neuropsychiatric disorders. Many forms of mental illness including autism and schizophrenia are associated with abnormal alterations in the limbic circuitry. Thus, our studies should also yield novel insights into the etiology and treatment of such disorders.